(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Long-term effects of noncontingent reinforcement on behavior maintained by automatic reinforcement"

LONG-TERM EFFECTS OF NONCONTINGENT REINFORCEMENT ON 
BEHAVIOR MAINTAINED BY AUTOMATIC REINFORCEMENT 



By 
JANA SEITER LINDBERG 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE 

UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE 

REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY 

UNIVERSITY OF FLORIDA 

2000 



ACKNOWLEDGMENTS 

I would like to thank those who helped make this research possible. First, I thank 
Dr. Brian Iwata, my advisor and supervisory committee chair, for teaching me to evaluate 
and conduct behavioral research. His advice and support throughout this project have 
been invaluable. I also thank the other committee members--Drs. Jennifer Asmus, Marc 
Branch, Shari Ellis, and Timothy Vollmer--for their assistance. In addition, I thank my 
colleagues, Gregory Hartley and Eileen Roscoe, who acted as therapists for the 
experiments. Finally, I thank my husband, Brooks Lindberg, for believing in me and 
cheering me on. 



n 



TABLE OF CONTENTS 

page 

ACKNOWLEDGMENTS ii 

LIST OF FIGURES iv 

ABSTRACT v 

INTRODUCTION 1 

Automatic Reinforcement 1 

Assessment of Behaviors Maintained by Automatic Reinforcement 2 

Treatment of Behaviors Maintained by Automatic Reinforcement 5 

Noncontingent Reinforcement (NCR) 8 

Rationale for Current Investigation 19 

EXPERIMENT 1 22 

Methods and Results 23 

Participants and Setting 23 

Response Measurement and Reliability 24 

Phase 1: Preference Assessment 24 

Phase 2: Brief NCR Evaluation 25 

Phase 3: Extended NCR Evaluation 27 

Discussion 30 

EXPERIMENT 2 32 

Methods and Results 33 

Participants and Setting 33 

Response Measurement and Reliability 33 

Phase 1: Functional Analysis 34 

Phase 2: Preference Assessment 35 

Phase 3: Brief NCR Evaluation 39 

Phase 4: Extended NCR Evaluation 41 

Phase 5: Naturalistic NCR Evaluation 43 

Discussion 44 

GENERAL DISCUSSION 47 

REFERENCES 57 

BIOGRAPHICAL SKETCH 61 

• • ■ 

in 












LIST OF FIGURES 

Figure ESge 

1 Leisure item preference assessment results for Matthew and Angela 26 

2 Percentage of intervals of item contact for Matthew and Angela across 1 0-min 
baseline and analog NCR sessions 28 

3 Percentage of intervals of item contact for Matthew and Angela across 1 20-min 
NCR constant and NCR varied sessions 29 

4 Rates of SIB for Laura and Robert across functional analysis conditions 36 

5 Leisure item preference assessment results for Laura and Robert 38 

6 Rates of SIB for Laura and Robert across 1 0-min baseline and NCR sessions 40 

7 Rates of SIB for Laura and Robert across 1 20-min NCR constant and NCR varied 
sessions 42 

8 Rates of SIB for Laura during 10-min observation periods at home 45 



IV 



Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 

LONG-TERM EFFECTS OF NONCONTINGENT REINFORCEMENT ON 
BEHAVIOR MAINTAINED BY AUTOMATIC REINFORCEMENT 

By 

Jana Seiter Lindberg 

May 2000 

Chairman: Brian A. Iwata 
Major Department: Psychology 

The purpose of the current investigation was to evaluate the long-term effects of 

noncontingent reinforcement (NCR) on behavior problems maintained by automatic 

reinforcement. In the first experiment, treatment effects were examined by studying 

behavior that can be considered analogous to the response options available when NCR is 

used to treat self-injurious behavior (SIB) maintained by automatic reinforcement: 

Manipulation of a low- or medium-preference leisure item was likened to engaging in 

SIB maintained by automatic reinforcement, and manipulation of a high-preference item 

was likened to manipulating a competing leisure item available during NCR. Two 

individuals participated. An assessment was conducted to identify low-, medium-, and 

high-preference leisure items. The effects of NCR were then evaluated during 10-min 

and 120-min sessions. 



In the second experiment, the effects of NCR were evaluated on the SIB of two 
individuals. First, functional analyses were conducted to determine that the participants' 
SIB was not maintained by social reinforcement. Assessments were also conducted to 
identify leisure items associated with long durations of manipulation and low levels of 
SIB. The effects of providing continuous access to a highly preferred leisure item were 
then assessed during 10-min and 120-min sessions. Varied reinforcers were subsequently 
delivered during 120-min sessions to determine if treatment effects might be extended. 
The effects of using NCR all day were also assessed over several months for one 
participant. 

The results of Experiments 1 and 2 demonstrated that reinforcers obtained 
through item manipulation can compete with the reinforcers obtained automatically by 
engaging in SIB during brief NCR sessions. However, data from the 120-min sessions 
indicated that NCR may lose its effectiveness when used for long periods of time. 
Providing varied high-preference leisure items may extended the usefulness of NCR for 
some individuals. When NCR was implemented all day, its therapeutic effects were 
shown to last over several months. Thus, NCR may reduce some individuals' SIB over 
long periods, but additional interventions may be necessary for others. 



VI 



INTRODUCTION 

The purpose of the current investigation was to evaluate the long-term effects of 
noncontingent reinforcement (NCR) on behavior problems maintained by automatic 
reinforcement. To provide the reader with a background to the current research, this 
paper will (a) explain the concept of automatic reinforcement, (b) describe how behavior 
is shown to be maintained by automatic reinforcement, (c) summarize five treatment 
strategies for behavior problems maintained by automatic reinforcement, (d) review the 
history of noncontingent reinforcement as a treatment for behavior disorders, and (e) 
explain the rationale for conducting the current study. 

Automatic Reinforcement 

Many persons with developmental disabilities engage in repetitive behaviors that 
persist in the absence of social reinforcement. These behaviors are said to be maintained 
by automatic reinforcement to the extent that they directly produce their own reinforcing 
consequences. The concept of automatic reinforcement was first introduced by Skinner 
(1953), who used the term to describe reinforcement that does not involve mediation by 
another person. Vaughn and Michael (1982) clarified and extended the concept of 
automatic reinforcement when they explained that automatic reinforcement "is a 
'natural 'result of behavior when it operates upon the behaver's own body or the 



2 
surrounding world. In general, the reinforcement may be conditioned or unconditioned, 

positive or negative" (p. 219). 

Assessment of Behaviors Maintained by Automatic Reinforcement 
Identifying the influence of automatic reinforcement on behavior may be difficult 
because reinforcement is produced directly by the behavior and often cannot be 
manipulated independently of the behavior (Vollmer, 1994). Nevertheless, Shore and 
Iwata (1999) suggested four strategies for assessing behavior that is suspected of being 
maintained by automatic reinforcement. First, the behavior should be shown not to be 
differentially sensitive to social consequences, which is best demonstrated by conducting 
a functional analysis (Iwata, Dorsey, Slifer, Bauman, & Richman, 1982/1994), in which 
an individual is exposed to test and control conditions. If the target behavior is not 
differentially high in one or more of the test conditions for social reinforcement (i.e., 
attention or escape), results would suggest that the behavior is insensitive to social 
contingencies. In other words, if behavior occurs at comparable across levels in all 
conditions (e.g., Vollmer, Marcus, & LeBlanc, 1994) or occurs at lower levels in the test 
conditions for social reinforcement than in the test condition for automatic reinforcement 
(e.g., Kennedy & Souza, 1995), it is unlikely that the behavior is maintained by social 
reinforcement. 

The second step is to verify that the behavior persists in the absence of social 
reinforcement, which is most often accomplished by observing the individual while 
alone. Vollmer, Marcus, Ringdahl, and Roane (1995b) have suggested that, at the 
conclusion of a functional analysis, an individual might be repeatedly observed while 
alone to verify that the behavior continues to occur when the test conditions for social 



3 
reinforcement are no longer being conducted. Persistence of behavior in the absence of 

social interaction rules out the possibility that undifferentiated functional analysis results 

were caused by intermittent social reinforcement, idiosyncratic reinforcers, multiple 

control, adjunctive schedules, or sequence effects. 

Third, if the behavior operates on the external environment and produces an 
observable response product, the consequence may be subject to manipulation. A 
functional relation is demonstrated by showing that the behavior occurs when it produces 
the observable consequence but does not occur when the consequence is prevented. For 
example, Rincover (1978) controlled one individual's level of plate spinning by 
attenuating the auditory stimulus produced by the behavior. The individual spun the plate 
more often when the table was left bare; plate spinning decreased when carpet was placed 
on the table. 

If the behavior does not produce an observable consequence or if it produces a 
consequence that is difficult to manipulate, then substitutable reinforcers may be 
identified that compete with the behavior. Substitutable reinforcers are "different 
reinforcers that are nevertheless interchangeable under certain conditions" (Shore, Iwata, 
DeLeon, Kahng, & Smith, 1997, p. 130). Finding substitutable reinforcers may help 
identify the nature of the automatic reinforcer that maintains the target behavior because 
the reinforcers may share common properties. 

It has also been suggested (Vollmer, 1994) that if behavior occurs at high rates in 
all functional analysis conditions, a medical examination should be conducted to 
determine if the behavior is maintained by automatic negative reinforcement. Behavior 
maintained by automatic negative reinforcement attenuates or delays aversive 



4 
physiological stimulation. For example, scratching may reduce irritation caused by skin 

conditions, and head hitting may attenuate pain caused from a headache or ear infection. 

Behavior maintained by automatic negative reinforcement is likely to occur whenever the 

individual experiences pain or discomfort. Thus, if the behavior occurs across all 

conditions or if the behavior occurs in a cyclical pattern, the behavior may be maintained 

by automatic negative reinforcement, and the source of discomfort should be identified 

and eliminated whenever possible. 

These strategies have been used to assess a number of behaviors exhibited by 

individuals with developmental disabilities. Researchers have found that automatic 

reinforcement has maintained behaviors such as eye poking (Kennedy & Souza, 1995; 

Lalli, Livezey, & Kates, 1996), face slapping (Van Houten, 1993), hand mouthing (Goh 

et al., 1995; Mazaleski, Iwata, Rodgers, Vollmer, & Zarcone, 1994), hand biting 

(Ringdahl, Vollmer, Marcus, & Roane, 1997; Vollmer et al., 1994), hair pulling (Rapp, 

Miltenberger, Galensky, Ellingson, & Long, 1999), head banging (Ringdahl et al., 1997; 

Vollmer et al., 1994), pica (Piazza et al., 1998), property destruction (Fisher, Lindauer, 

Alterson, & Thompson, 1998), skin picking and rubbing (Roscoe, Iwata, & Goh, 1998; 

Shore et al., 1997), and stereotypy (Fisher et al., 1998). In a large-scale experimental 

study (Iwata et al., 1994), functional analysis data from 152 individuals who engaged in 

various forms of self-injurious behavior (SIB) indicated that automatic reinforcement 

accounted for 25.7% of the cases. Thus, automatic reinforcement appears to be a 

significant source of reinforcement for many behavior problems. 






5 
Treatment of Behaviors Maintained by Automatic Reinforcement 

There are several reinforcement-based treatments that decrease behaviors 
maintained by automatic reinforcement. These treatments can be categorized into three 
different approaches: altering establishing operations, limiting reinforcement for the 
behavior, and strengthening competing behaviors. 

Michael (1993) described an establishing operation as "an environmental event, 
operation, or stimulus condition that affects an organism by momentarily altering (a) the 
reinforcing effectiveness of other events and (b) the frequency of occurrence of that part 
of the organism's repertoire relevant to those events as consequences" (p. 192). The most 
common establishing operations for behavior maintained by positive and negative 
reinforcement are deprivation and aversive stimulation, respectively. Thus, behavior 
maintained by positive reinforcement may be less likely to occur if the individual is 
satiated to the reinforcer maintaining the behavior; behavior maintained by negative 
reinforcement is unlikely to occur when the aversive stimulation is no longer present. 

Another approach to treating behavior maintained by automatic reinforcement is 
to limit reinforcement for the target behavior. This may be accomplished by 
extinguishing the behavior or by increasing the effort required to engage in the response. 
Extinction requires the discontinuation or attenuation of the reinforcement maintaining 
the behavior. Rincover (1978) reported using extinction to decrease the stereotypic 
behaviors of three individuals. For example, the behavior of one individual who twirled a 
plate on a table decreased when the sound produced by the behavior was attenuated by 
carpeting the table. Vibrators were attached to the backs of two other individuals' hands 
to attenuate the stimulation produced by finger flapping (one child) and object twirling 



6 
(second child). In other studies, investigators have used protective equipment to reduce 

the stimulation produced by behavior (Dorsey, Iwata, Reid, & Davis, 1982; Parrish, 

Aguerrevere, Dorsey, & Iwata, 1980). 

Zhou, Goff, and Iwata (2000) demonstrated that increasing response effort is 
another viable way to limit reinforcement for behaviors maintained by automatic 
reinforcement. Four individuals who engaged in high levels of hand mouthing even 
when they had access to preferred leisure items participated. When they wore soft, 
flexible sleeves that increased resistance for elbow flexion, the participants engaged in 
lower levels of hand mouthing. Interestingly, application of the device did not interfere 
with the high levels of object manipulation exhibited by two individuals and actually 
increased object manipulation by the remaining two participants. 

Finally, behavior maintained by automatic reinforcement may be decreased by 
strengthening competing behaviors through differential reinforcement. This approach 
usually involves delivering reinforcement contingent on the occurrence of appropriate 
behavior and withholding reinforcement contingent on the occurrence of problem 
behavior. For example, Wacker et al. (1990) used differential reinforcement of 
alternative behavior (DRA) to decrease one individual's stereotypy. The individual was 
given access to a rocking chair contingent on emitting an appropriate communicative 
response. Favell, McGinsey, and Schell (1982) found that the SIB of 6 individuals 
decreased when participants had access to alternative activities. SIB decreased even 
further when object manipulation was reinforced. Cowdery, Iwata, and Pace (1990) used 
differential reinforcement of other behavior (DRO) to decrease the severe SIB of one 



7 
individual; the individual received access to video games and other preferred activities 

contingent on the nonoccurrence of SIB. 

In summary, treatments based on altering establishing operations, limiting 
reinforcement for the behavior, and strengthening competing behaviors have been shown 
to reduce problem behavior maintained by automatic reinforcement. However, the latter 
two approaches have limitations that may make them less attractive than the first 
approach. In order to implement extinction, the source of reinforcement must be 
identified and controlled. This is difficult because behavior may produce multiple forms 
of stimulation, and even if identified, intrusive measures (i.e., protective equipment) may 
be required to limit reinforcement of the response. Increasing response effort may also 
be problematic because some form of equipment is usually required, and in some cases 
the response topography does not easily lend itself to increasing the effort required to 
engage in the response. For example, it would be difficult to increase the effort required 
to engage in problem behavior such as head banging, echolalia, or spitting. Treatment 
approaches based on strengthening alternative responses may be unsuccessful if highly 
preferred reinforcers cannot be identified: Reinforcers that are more valuable than the 
automatic reinforcer obtained by engaging in problem behavior are required for 
differential reinforcement procedures because it is usually not possible to put the target 
behavior on extinction. Thus, a typical differential reinforcement procedure is much like 
a concurrent reinforcement schedule in which the individual receives one reinforcer for 
the problem behavior or another reinforcer for the alternative behavior. 

Treatment approaches based on altering establishing operations are advantageous 
because they are usually easy for caregivers to implement and may be effective even 



8 
when the target behavior continues to be reinforced. The latter advantage is especially 

important because the reinforcers maintaining the behavior may be impossible to identify 

and control. One of the most common methods of altering establishing operations is 

noncontingent reinforcement (NCR). 

Noncontingent Reinforcement (NCR) 

NCR involves the delivery of a reinforcer according to a schedule that is 
independent of the occurrence of specific behavior. The effects of NCR were 
demonstrated in an early study (Rescorla & Skucy, 1969) in which NCR was compared 
to extinction. During baseline, bar-pressing by rats was reinforced with food pellets 
according to a variable-interval (VI) schedule. Subsequently, some of the rats were 
exposed to an extinction condition in which no food was delivered, whereas others were 
exposed to an NCR procedure in which food pellets were delivered on a variable-time 
(VT) schedule in which the rate of reinforcement was yoked to that received by a third 
group of rats that continued to receive reinforcement for bar-pressing on a VI schedule. 
Both extinction and NCR decreased bar-pressing by rats, but response suppression Was 
greater with extinction. 

The term "noncontingent reinforcement" has been the subject of recent debate 
(Poling & Normand, 1999; Vollmer, 1999). The term has been used to describe a 
procedure in which a stimulus with known reinforcing properties is delivered according 
to response-independent or time-based schedule (Vollmer, Iwata, Zarcone, Smith, & 
Mazaleski, 1993). However, Poling and Normand criticized the practice of using the 
term "reinforcement" to describe a situation in which the delivery of a stimulus does not 
result in an increase in any behavior that is measured. The authors argued that the 



9 
function of a stimulus can vary depending on the context: A stimulus may function as a 

reinforcer in one setting when it is delivered on one schedule, but it may not increase 

behavior in another setting or when delivered according to another schedule. Thus, the 

authors concluded that "the fact that a stimulus is a positive reinforcer in one context 

does not justify terming it a positive reinforcer in a context in which its delivery reduces 

responding" (p. 237). 

Vollmer (1999) agreed that the term noncontingent reinforcement is problematic 
and encouraged researchers to describe their procedures in operational terms such as 
"fixed-time attention" or "fixed-time escape." However, he proposed that the term 
noncontingent reinforcement has served a good purpose because it seems to have led 
behavior analysts to recognize the potential of NCR as a general class of procedures 
rather than as a unique application of a time-based schedule. Vollmer explained that, 
"Calling the procedure NCR gave it status as a treatment package on par with DRO 
[differential reinforcement of other behavior], insofar as both names describe a general 
procedure that is not limited to any particular stimulus or event" (p. 240). In addition, 
NCR has the advantage of capturing an important aspect of the procedure that terms like 
"fixed-time attention" or "response-independent escape" do not: The stimulus delivered 
has been shown to be a reinforcer for the participant and, in most cases, is the same 
stimulus that has been shown to maintain the target behavior. For these reasons, the term 
noncontingent reinforcement, or NCR, will be used to describe past research and the 
current investigation. 

NCR has been used to treat a number of behavior problems maintained by social 
reinforcement. Usually, when the target behavior is maintained by social reinforcement, 






10 
the reinforcer that maintains the target behavior is no longer delivered following the 

target behavior during NCR. Instead, the maintaining reinforcer is usually delivered 

according to a time-based schedule (i.e., independently of the target behavior). Thus, the 

maintaining reinforcer is still delivered, but the dependency between the target behavior 

and reinforcement is discontinued. 

An example of how NCR may be used as treatment for behavior maintained by 
social reinforcement was described by Mace and Lalli (1991), who found that one 
individual's bizarre vocalizations were maintained by social positive reinforcement in the 
form of attention. The authors developed two treatments to decrease the vocalizations: 
NCR and communication training. In the NCR condition, attention was delivered 
according to a VT schedule. During communication training, the individual was first 
taught to initiate a conversation and was then taught to expand the conversation by asking 
questions or offering contextually appropriate information. Both interventions effectively 
decreased bizarre vocalizations. 

Vollmer et al. (1993) compared the effects of two interventions, DRO and NCR, 
on the SIB of three individuals. The authors conducted a functional analysis of the 
participants' SIB and found that each subject's target behavior was maintained by social 
positive reinforcement in the form of attention. The participants were then exposed to 
DRO and NCR in either multielement or reversal designs. During DRO sessions, a 
therapist delivered attention to the participant contingent on the absence of SIB for a 
specified interval. During NCR sessions, a therapist delivered attention to the participant 
based on an FT schedule. DRO and NCR produced comparable decreases in SIB; 
however, the authors recommended NCR over DRO for several reasons: (a) extinction- 



11 

induced behavior was attenuated for 2 of the 3 participants during NCR, (b) the 
participants received more reinforcers during NCR, and (c) NCR was easier to implement 
because caregivers were not required to constantly monitor the participant and reset the 
DRO interval each time the target behavior occurred. 

Hagopian, Fisher, and Legacy (1994) examined the effects of dense versus lean 
schedules of NCR on the destructive behaviors of four children. The authors first 
conducted a functional analysis and found that the children's target behaviors were 
maintained by attention. The authors then compared a dense NCR schedule, in which the 
participant received attention continuously, to a lean schedule, in which the individual 
received attention once every 5 min. Results showed that the dense NCR schedule 
produced immediate and dramatic reductions in destructive behavior in all four 
participants, whereas the lean schedule was much less effective. Following this 
comparison, the dense NCR schedule was successfully thinned from FT 10 s to FT 5 min. 

Fischer, Iwata, and Mazaleski (1997) used NCR in the absence of extinction to 
decrease two participants' SIB. The authors studied the extent to which noncontingent 
delivery of arbitrary reinforcers (i.e., reinforcers that are irrelevant to behavioral 
maintenance) would decrease problem behaviors maintained by social reinforcement. 
Results of a functional analysis demonstrated that one participant's SIB was maintained 
by attention, and the other's SIB was maintained by access to a preferred clothing item. 
An additional assessment also demonstrated that preferred food items did not maintain 
either participant's SIB. During NCR, the target behaviors continued to produce then- 
maintaining reinforcers, while preferred food items were delivered on an FT schedule. 



12 
Even though the target behavior continued to be reinforced, the behavior decreased 

during NCR. 

Vollmer, Marcus, and Ringdahl (1995a) extended research on NCR by using the 
procedure to decrease behavior maintained by social negative reinforcement in the form 
of escape from instructional activities. Two young men who engaged in SIB participated 
in the study. A functional analysis was conducted, and the results suggested that both 
participants' target behavior was maintained by escape. During NCR, escape from a 
learning task was provided on an FT schedule. Initially, escape was provided 
continuously, but the schedule was gradually thinned to FT 2.5 min for one participant 
and to FT 1 min for the other. The authors found that NCR was an effective treatment 
for behavior maintained by escape. 

NCR has also been used as a treatment for behavior problems maintained by 
automatic reinforcement. In these cases, it is usually not possible to deliver the same 
reinforcer that maintains the target behavior because the target behavior produces a 
product that cannot be manipulated or delivered independently of the behavior. 
Nonetheless, Shore and Iwata (1999) have observed that providing "access to similar 
(substitutable) but different reinforcement might abolish the establishing effects of either 
deprivation or aversive stimulation" (p. 132). In most cases, NCR procedures for 
behavior maintained by automatic reinforcement have involved providing continuous 
access to leisure items. In some studies, authors have reported providing continuous 
access to food (Favell et al., 1982; Piazza et al., 1998) or to some other stimulus (e.g., 
Bailey & Meyerson, 1970). 



13 
In one study (Bailey & Meyerson, 1970), a child's severe SIB decreased when a 

vibrator was activated that made the mattress of his crib shake. Providing the vibration 

noncontingently during 10-min periods was more effective than providing vibration for 6 

s contingent on lever-pressing. One interesting aspect of this study was that, during 

NCR, the participant was not required to do anything to obtain the reinforcer (vibration). 

In a number of other studies, NCR involved providing leisure items that the 
participants could manipulate in order to access reinforcement. For example, in one 
study (Lockwood & Bourland, 1 982), continuous access to leisure items was shown to 
decrease hand biting in one participant and arm biting and face slapping in another. 
Leisure items, such as colorful rubber and soft plastic toys, were selected that were 
thought to provide stimulation similar to that provided by the target behaviors. 
Incidentally, the authors found that noncontingent access to the leisure items was most 
effective when the leisure items were attached to the participants' wheelchairs rather than 
left loose on the participants' laps. 

Favell et al. (1982) decreased the SIB of 6 individuals by providing alternate 
activities that produced sensory stimulation similar to that apparently obtained from 
engaging in the target behaviors. One individual who chewed and sucked on his hands 
was given large, soft items that could be mouthed. Two individuals who engaged in eye- 
poking were given leisure items with striking visual properties. Three individuals who 
engaged in pica were given popcorn and leisure items that required hand manipulation. 
SIB decreased substantially when participants had access to these items and decreased 
even further when object manipulation was reinforced. 









14 
Vollmer et al. (1994) demonstrated the importance of providing preferred items 

during NCR to obtain decreases in SIB. Functional analyses were conducted for three 

participants, and results suggested that all three participants' target behavior was 

maintained by automatic reinforcement. A preference assessment was also conducted to 

identify preferred leisure or food items for the participants. 

The authors then compared the effects of providing continuous, noncontingent 
access to preferred versus non-preferred items and found that noncontingent access to 
preferred stimuli produced decreases in SIB for all participants, whereas noncontingent 
access to non-preferred stimuli had little effect on behavior. To obtain further decreases 
in SIB, the authors reinforced the object manipulation of two participants, and manually 
restrained one of these two participants for 5 s following occurrences of SIB. The 
authors also found that NCR began to lose its effects for one participant until another 
preference assessment was conducted and new items were made available during NCR. 
In the final phase, two participants' families were trained to implement treatment 
packages that included the NCR procedure for one hour per day. The authors visited the 
participants' homes once per week for 7 weeks to evaluate the effectiveness of the 
intervention package. Follow-up data were also collected for one participant one and five 
months later. Results indicated that the intervention remained effective for both 
participants when implemented under naturalistic conditions. 

Shore et al. (1997) suggested that the results of many studies on NCR as a 
treatment for behavior problems maintained by automatic reinforcement may be analyzed 
in terms of reinforcer substitutability. The effects of NCR depend on the relation 
between the arbitrary reinforcer that is delivered noncontingently and the reinforcer 






15 
obtained through engaging in the target behavior. Three relations are possible: 

complementarity, substitutability, and independence. "Substitutability describes a 

continuum of interactions between concurrently available reinforcers. At one end of the 

continuum are complementary reinforcers, for which increased consumption of one 

alternative results in increased consumption of its complement. ... At the other end of the 

continuum are substitutable reinforcers, for which an increase in consumption of one 

alternative results in decreased consumption of its substitute.. . . In the middle of the 

continuum, reinforcers are independent: Consumption of one has minimal effect of 

consumption of another" (p. 23). Therefore, if the reinforcer delivered noncontingently 

and the reinforcer obtained through the target behavior are complementary, the target 

behavior would increase. If the two are substitutable, NCR would produce decreases in 

the target behavior, and if the two are independent, the treatment may have no effect. 

Shore et al. (1997) examined the relation between the automatic reinforcement 
obtained through SIB and the reinforcement obtained through leisure item manipulation. 
They first conducted a functional analysis of the SIB of three individuals to determine 
that the behavior was maintained by automatic reinforcement. Probes were then 
conducted to identify leisure items that appeared to produce reinforcers substitutable for 
SIB. The authors then evaluated the effects of NCR using the leisure items identified in 
the probes. NCR produced large decreases in SIB for all three participants. 

Shore et al. (1997) conducted two additional experiments to investigate further the 
extent to which arbitrary reinforcers would compete with those obtained through SIB. In 
one experiment, the authors evaluated the effects of delivering the leisure items according 
to various DRO schedules but failed to observe clinically significant reductions in SIB 



16 
for any of the 3 participants. In the final experiment, the authors altered the effort 

required to manipulate preferred leisure items while keeping constant the effort required 

to engage in SIB. Response effort of item manipulation was altered by anchoring objects 

to a lap tray with a string and varying the length of the string. When the string was at full 

length, the individual could manipulate the object while seated in an upright position. As 

the string was shortened, the individual had to bend over further and further to 

manipulate the item. The authors found that when the string was at full length, the 

participants engaged in high levels of object manipulation and low levels of SIB. As the 

string was shortened, object manipulation decreased and SIB increased, until a switch in 

preference between object manipulation and SIB was observed. These results illustrate 

that the relationship between two reinforcers (in this case, the reinforcer maintaining SIB 

and the reinforcer obtained through object manipulation) depends on the context in which 

they are presented. 

Two recent studies (Goh et al., 1995; Piazza et al., 1998) have indirectly 

investigated the nature of the automatic reinforcer that maintains behavior by finding 

substitutable reinforcers and identifying the specific aspects of reinforcement that may 

make them substitutable. Goh et al. conducted functional analyses of the hand mouthing 

of 12 individuals. Results from 10 of the 12 individuals' assessments suggested that the 

target behavior was maintained by automatic reinforcement; results for the remaining two 

individuals suggested that the behavior was maintained by social positive reinforcement. 

Four of the individuals whose hand mouthing appeared to be maintained by automatic 

reinforcement participated in a second experiment to identify the possible specific 

sources of reinforcement for the target behavior. The participants were given access to 



17 
an object they had been observed to manipulate outside of sessions while the 

experimenters measured the participants' levels of hand-mouth, hand-item, and mouth- 
item contact. Results indicated that they preferred hand-item contact. In the next 
experiment, the experimenters measured hand-mouth, hand-item, and mouth-item contact 
of five individuals across a variety of leisure items. All five individuals engaged in 
greater levels of hand-item contact than hand-mouth or mouth-item contact. From these 
data, the authors concluded that hand stimulation may have been the more important 
reinforcing aspect of hand mouthing for these individuals. 

Piazza et al. (1998) conducted functional analyses of the pica of three individuals. 
Results of the assessment suggested that the pica of one participant was maintained by 
automatic reinforcement and the pica of the other two participants appeared to be 
maintained by both automatic and social reinforcement. Next, the authors conducted 
preference assessments to find stimuli that would compete with pica. They assessed 
stimuli that produced oral stimulation (matched stimuli) and stimuli that produced other 
types of stimulation (unmatched stimuli). The authors hypothesized that pica was 
maintained by the oral stimulation it produced. To test this hypothesis, the authors 
presented preferred matched stimuli in one condition and preferred unmatched stimuli in 
another condition. Pica occurred at lower levels in the matched stimuli condition for 2 of 
the 3 participants. 

The two individuals who exhibited lower levels of pica during the matched 
stimuli condition participated in a more detailed assessment to identify the aspect of oral 
stimulation that served as reinforcement. The authors hypothesized that firmness (rather 
than taste or other aspects of texture) was an important component of pica for the 



18 
individuals and assessed items with varying degrees of firmness (e.g., rice cakes, carrot 

sticks, tofu, gelatin, etc.). When the participants had access to firm items, they engaged 

in lower levels of pica than when they had access to soft items. These results supported 

the hypothesis that firmness was an important aspect of reinforcement for pica for these 

individuals. 

Roscoe et al. (1998) compared the effects of NCR to those of protective 
equipment on the SIB of three individuals. They first conducted a functional analysis and 
found that all three individuals' target behaviors were maintained by automatic 
reinforcement. Next, the authors conducted probes to identify a leisure item for each 
individual that would effectively compete with SIB. The leisure items selected were a 
massager, a plastic ring, and a small musical keyboard. Additional probes were 
conducted to identify the least intrusive form of protective equipment that effectively 
decreased each participant's form of SIB. Foam sleeves were selected for one individual 
who rubbed and hit his arms against hard stationary surfaces, boxing gloves were selected 
for the participant who engaged in hand mouthing, and latex gloves were selected for the 
participant who picked and rubbed her skin. The authors conducted alone baseline 
sessions and then compared the effects of NCR and protective-equipment sessions in 
multielement designs. Both procedures effectively decreased SIB, but the authors 
recommended using NCR because (a) it produced slightly more rapid or more complete 
suppression, (b) it required little effort to implement, and (c) it occasioned appropriate 
alternative behaviors. 

The above research has demonstrated that NCR can effectively decrease behavior 
maintained by automatic reinforcement. Recent studies have highlighted the importance 






19 

of selecting highly preferred reinforcers to be delivered during NCR (Piazza et al., 1998; 

Vollmer et al., 1994). Results from other studies have suggested that it may be beneficial 
to present stimuli that appear to provide stimulation similar to that obtained through the 
target behavior (Favell et al., 1982; Goh et al., 1995; Lockwood & Bourland, 1982; 
Piazza et al., 1998; Shore et al., 1997). However, even when highly preferred or 
presumably matched reinforcers are delivered, sometimes NCR is not effective and 
additional procedures are necessary (Lindberg, Iwata, & Kahng, 1999; Ringdahl et al., 
1997; Vollmer etal., 1994). 

Rationale for Current Investigation 

Research on noncontingent reinforcement has demonstrated that the intervention 
can be used to reduce a variety of behavior problems. However, these findings are based 
on evaluations of NCR during brief experimental sessions, ranging in duration from 5-15 
min. It is unclear if the intervention would remain effective when evaluated over longer 
periods of time. This is particularly questionable when NCR is used to decrease behavior 
maintained by automatic reinforcement. 

To understand why the effects of NCR are less likely to be maintained over time 
when the procedure is used to treat behavior maintained by automatic reinforcement than 
when NCR is used for behavior maintained by social reinforcement, it is important to 
understand several key differences between the two procedures: the inclusion or 
exclusion of an extinction component, the relationship of the stimulus delivered 
noncontingently to the stimulus maintaining the target behavior, and the effects of 
extended exposure to noncontingent reinforcement. 






20 
A behavior maintained by social reinforcement persists during baseline because 

the maintaining reinforcer is delivered by another individual. When NCR is 

implemented, the maintaining reinforcer is usually no longer delivered contingent on the 

target response. By contrast, a behavior maintained by automatic reinforcement persists 

during baseline because it directly produces the reinforcer. Because the target behavior is 

maintained by automatic reinforcement, occurrences of the behavior continue to produce 

reinforcement during treatment. Thus, in the first case, NCR can include an extinction 

component, whereas in the second case, it usually cannot. 

The second key difference between NCR for behavior maintained by social 
reinforcement and NCR for behavior maintained by automatic reinforcement involves the 
relationship of the stimulus delivered noncontingently to the stimulus maintaining the 
target behavior. When NCR is used to decrease a behavior maintained by social 
reinforcement, the same reinforcer that maintains the target behavior is usually delivered 
during treatment. By contrast, when NCR is used to treat behavior maintained by 
automatic reinforcement, an arbitrary reinforcer is delivered that competes with the 
maintaining reinforcer. 

Given the above considerations, satiation to the reinforcer delivered during NCR 
would have different effects depending on whether the intervention is used to treat a 
behavior maintained by social or automatic reinforcement. When the intervention is used 
for behavior maintained by social reinforcement, satiation to the reinforcer being 
delivered noncontingently would not cause the target behavior to increase because the 
reinforcer maintaining the behavior is being delivered noncontingently. By contrast, 
when NCR is implemented to treat behavior maintained by automatic reinforcement, 



21 
satiation to the noncontingent reinforcer may lead to an increase in the target response 

because the reinforcer delivered noncontingently is an arbitrary reinforcer. Satiation to 
the arbitrary reinforcer may not alter the establishing operation of the target behavior; 
once consumption of the noncontingent reinforcer decreases, the individual may be 
motivated to engage in the target behavior. These satiation effects, if they occur, are 
more likely to be observed when NCR is evaluated over longer periods of time. 

The purpose of the current investigation was to determine whether NCR involving 
the delivery of arbitrary reinforcers leads to reemergence of problem behavior due to 
satiation. A secondary purpose was to determine if satiation could be mitigated through 
the use of multiple reinforcers. Two experiments were conducted. In the first, the long- 
term effects of NCR were investigated in an analog situation. To accomplish this goal, 
two individuals' patterns of leisure item manipulation were investigated under a variety 
of conditions. In the second experiment, the long-term effects of NCR on SIB were 
investigated. Two individuals who engaged in SIB maintained by automatic 
reinforcement participated. 



EXPERIMENT 1 

The purpose of this experiment was to evaluate the long-term effects of NCR by 
observing the relationship between behavior that was hypothesized to be analogous to 
target and alternative behaviors when NCR is used to treat behavior maintained by 
automatic reinforcement. Manipulation of a low- or moderate-preference leisure item 
was considered analogous to engaging in SIB, whereas manipulation of a high-preference 
leisure item was likened to consuming the competing reinforcer (i.e., playing with a toy 
or eating food) that is delivered noncontingently during treatment. Engaging in SIB was 
compared to manipulating a lesser-preferred leisure item, and consuming the arbitrary 
(noncontingent) reinforcer was compared to manipulating a more-preferred item because, 
when NCR is effective, the participant allocates more time to consuming the arbitrary 
reinforcer and less time engaging in SIB. We selected leisure items that were 
differentially preferred by the participant so as to increase the likelihood that the 
individual would manipulate one item more than the other during NCR. 

During baseline, one leisure item was available. This was considered comparable 

to the conditions that exist when levels of SIB maintained by automatic reinforcement are 

measured during typical baseline sessions: The individual is placed in a situation in 

which little reinforcement is available except by engaging in SIB. During the analogue 

NCR sessions, two leisure items were available concurrently. This situation was likened 

to the conditions of a treatment session in that, during NCR, the individual may access 

22 



23 
reinforcement by engaging in SIB or by consuming the reinforcer that is delivered 

noncontingently. 

The study was conducted in 3 phases. In the first phase, participants' preferences 

for a number of leisure items were assessed. In the second phase, participants' levels of 

object manipulation with one or two items were recorded during brief sessions. In the 

third phase, levels of object manipulation were recorded during extended sessions. 

Methods and Results 

Participants and Setting 

Two individuals enrolled in a sheltered workshop program for persons with 
developmental disabilities participated. Matthew was a 32-yr-old man diagnosed with 
mental retardation. He was ambulatory, could follow multi-step instructions, and 
communicated vocally. Angela was a 38-yr-old woman also diagnosed with mental 
retardation. She was ambulatory and could follow simple one- and two-step instructions. 
Angela could communicate vocally but, due to articulation problems, she often 
communicated with gestures and a few manual signs. Neither Matthew nor Angela 
engaged in any behavior problems. 

All sessions were conducted on the grounds of the sheltered workshop. The 
sessions for one participant were conducted in an area that had a small table and a few 
chairs, which was partitioned off from the main workshop area. The second participant's 
sessions were conducted in a conference room with a large table and several chairs. 






24 
Response Measurement and Reliability 

The dependent variable was object manipulation, which was defined as physical 
hand contact with (e.g., touching, holding) a leisure item. Data were collected on 
handheld computers during continuous 10-s intervals and were summarized as the 
percentage of intervals during which responding occurred. Interobserver agreement was 
assessed by having a second observer independently collect data during 38.5% and 47.0% 
of Matthew's and Angela's sessions, respectively. Observers' records were compared on 
an interval-by- interval basis. An interval was considered an agreement if both observers 
scored either the presence or absence of behavior. Agreement coefficients were 
calculated by dividing the number of intervals containing agreements by the total number 
of intervals and multiplying by 100%. Mean agreement scores were 99.0% (range, 
96.7% to 100%) and 97.3% (range, 74.4% to 100%), respectively, for Matthew's and 
Angela's leisure item manipulation. 

Phase 1 : Preference Assessment 

The participants' preferences for a number of leisure items were assessed using 
procedures described by Fisher et al. (1992). Before the assessment, the therapist 
familiarized the participants with each leisure item. Then, each item was paired once 
with every other item, with the order of presentation determined randomly. On each trial, 
two leisure items were placed next to each other and approximately 30 cm in front of the 
participant. An approach response to one item produced 30-s access to that item while 
the other item was removed. Attempts to approach both items were blocked. If neither 
stimulus was approached within 5 s, the therapist prompted the participant to sample each 



25 
item and then repeated the trial. If the participant did not approach either item when the 

trial was repeated, the therapist removed both items and initiated a new trial. 

Figure 1 shows the percentage of trials on which each item was selected during 
the leisure item assessment. The items are listed from left to right in descending order 
based on the number of trials on which they were selected. Matthew selected the beads 
and string at every opportunity, and the sports magazine only once. The beads and string 
were selected as the high-preference item for inclusion in Phase 2. The sports magazine 
was selected as the low-preference item because it was clearly less preferred than the 
beads and string, but it was still selected on at least some of the trials. Angela selected 
the coloring book and crayons most often, and the mirror least often. The balloon was 
selected on 44.4% of the trials. The coloring book and crayons were selected as the high- 
preference item; the balloon was selected as the low-preference. The puzzle and the 
beads and string were selected as Angela's additional high-preference items for inclusion 
in Phase 3. 

Phase 2: Brief NCR Evaluation 

In the second phase, the effects of NCR were evaluated during 10-min sessions 
using a multiple baseline design. A reversal design was also used during Angela's 
assessment. During baseline, the participant had access to a low-preference item; during 
NCR, both a low-preference and a high-preference item were available. At the beginning 
of each session, the appropriate number of items was placed on a table in front of the 
participant. The participant was not given any instructions regarding item contact. At the 
conclusion of the session, the participant was informed that the session was over, and the 



26 



B 
o 

13 

CO 

•c 
H 



0) 

OB 

CO 

d 
<u 
o 

B 



lOO-i 

75- 
50- 

25- 



HP 



Matthew 



()■ 



CO 

o 

00 



8 

CQ 
oo 

c 

•g 

o 
U 



i 

o 
To 

N 



LP 



8 
o 

ft 



l_ 


JS 


<u 


3 


O 


B, 


Q 


£3 


o 


PL, 


| 


o 




w 


<u 


S 


1 
1 


-g 
<3 


3 


J3 




o 




o 
3 


3 

£ 



T3 

C 

E 

"2 

o 



c 

'n 
a 
oo 

c3 



lOO-i 

75- 
50- 
25- 



0- 



HP 



Angela 



HP 



HP 



LP 



<u 

N 
N 



00 

a 
•S 

— ' 
CO 
~co 

I 

CQ 



8 

'B 
O 
o 

'go 

3 

2 



2 

S3 


CO 

6 

3 


c 

1 

3 


O 


04 


CQ 


CQ 





CO 

■*— » 

ja 
op 

'5 



o 
fc 



Stimuli 



Figure 1: Leisure item preference assessment results for Matthew and Angela 



27 
materials were removed from the table. The therapist did not praise or otherwise 

reinforce object manipulation. 

Figure 2 shows the percentage of intervals of item contact for Matthew (top 

panel) and Angela (bottom panel). When only the low-preference item was available, the 

participants manipulated the item at high levels throughout the baseline sessions. The 

mean level of item manipulation during baseline sessions was 96.0% (range, 88.3 to 

100%) and 74.2 (range, 43.1 to 100%) for Matthew and Angela, respectively. When both 

the low- and high-preference items were available, Matthew and Angela manipulated the 

high-preference item during an average of 98.6% (range, 93.1 to 100%) and 97.6% 

(range, 92.0 to 100%) of the intervals, respectively. Manipulation of the low-preference 

item quickly dropped to zero for both participants. 

Phase 3: Extended NCR Evaluation 

The effects of NCR were evaluated during 120-min sessions using a reversal 
design. During the first condition, NCR constant, the individual had access to the low- 
and high-preference leisure items. This condition was identical to the NCR condition in 
Phase 2. During the second condition, NCR varied, the individual had access to the low- 
preference item and to three different high-preference leisure items. During this 
condition, all of the items were placed on a table in front of the participant, who was 
allowed to manipulate any of the items throughout the session. 

Figure 3 shows percentage of intervals of item contact during 120-min sessions. 
Responding in each session is divided into 20-min blocks. When the low- and high- 
preference items were available, Matthew manipulated the high-preference item to the 
exclusion of the low-preference item. This pattern of results was observed over four 120- 



28 



lOO-i 



75- 



50- 



«9 25- 

E 

u 

40 
C ft-l 

i— i 0i 

■i-j 

G 

s 

g 

o 

100 n 

o 





<D 75- 



50- 



25- 



04 



Baseline 




\ 



LP 




r 
5 



Analog NCR 



/ 



HP 



Matthew 



BL 




Analog NCR 



Angela 



10 



15 



20 



Sessions 



Figure 2: Percentage of intervals of item contact for Matthew and Angela 
across 10-min baseline and analog NCR sessions 



100 -i 



75- 



50- 



25- 



13 

I 



J 3 

Oh 
V / 

i 

O 
U 

£ 

CD 



III 5 



LP 



Constant 



100- 



75- 



50- 



25- 



04 



29 

NCR Constant 
■ i n i 




\ 



10 



Varied 




Varied 
HP 



OOOOOOPOOOOO 



T 

20 



Matthew 



Constant Varied 




30 



Angela 



oooooo 



IT 
40 



cxxmt^Sd 



20-min Segments 



Figure 3: Percentage of intervals of item contact for Matthew and Angela 
across 120-min NCR constant and NCR varied sessions 



30 
min sessions. His mean level of object manipulation of the high-preference item was 

99.3% (range 94.0% to 100%). Because Matthew never manipulated the low-preference 

item during the 120-min NCR sessions, he was not given additional high-preference 

items. 

When the low- and high-preference items were made available to Angela during 

the 120-min sessions, she manipulated the high-preference item initially, but as the 

session progressed, she began manipulating the low-preference item more often. 

However, when she was given access to varied high-preference items, manipulation of 

the high-preference items remained high throughout the 120-min sessions (M = 96.4%; 

range, 83.9% to 100%), whereas manipulation of the low-preference item remained at or 

near zero (M = 0.1%; range, 0% to 2.0%). These results were replicated when the NCR 

constant and the NCR varied conditions were presented a second time in a reversal 

design. 

Discussion 

Results of this study indicated that during brief sessions, reinforcement obtained through 
manipulation of a high-preference item competed with that obtained through 
manipulation of a low-preference item. However, during 120-min sessions, one of the 
two participants (Angela) showed apparent satiation to the reinforcer obtained through 
manipulation of the high-preference item, as reflected by decreased levels of contact with 
the high-preference item and a subsequent increase in levels of contact with the low- 
preference item. The effects of satiation appear to have been ameliorated when varied 
high-preference items were available because contact with the high-preference items 



31 
remained high during extended sessions, whereas contact with the low-preference item 

remained low. 

These data suggest that when NCR is used to treat behaviors maintained by 

automatic reinforcement, the treatment may lose its effects for some individuals when it 

is implemented over extended periods of time, whereas the treatment may remain 

effective for others. In cases in which the effects of NCR are not maintained for long 

periods, identifying and providing multiple high-preference items may extend the 

usefulness of the intervention. 



EXPERIMENT 2 

Two distinct patterns of results were observed during the final phase of 
Experiment 1 . One pattern suggested that when NCR is used to treat behaviors 
maintained by automatic reinforcement, the treatment may lose its effects when 
implemented for extended periods. The second pattern suggested that NCR may remain 
effective over long periods without additional intervention. The purpose of Experiment 2 
was to determine if either of these two patterns would be observed when NCR was used 
to treat SIB maintained by automatic reinforcement. Specifically, we wanted to know if 
NCR would remain effective when the treatment was used during 2-hr sessions and when 
used all day. The secondary purpose was to determine if reinforcer variation would 
ameliorate the effects of satiation if NCR lost its effectiveness over time. 

Experiment 2 was conducted in 5 phases. First, a functional analysis was 
conducted of two individuals' SIB. Second, the participants' leisure item preferences 
were assessed to identify items that might compete with SIB when delivered 
noncontingently. Third, the effects of delivering the most highly preferred leisure item 
during brief NCR sessions were assessed. Fourth, the effects of NCR were assessed 
during extended sessions. Finally, NCR was conducted at one participant's home, and 
brief observations were conducted periodically to assess the long-term effects of NCR 
under naturalistic conditions 



32 



33 
Methods and Results 

Participants and Setting 

Two individuals living in a state residential facility for persons with 
developmental disabilities and who engaged in SIB participated. Laura was a 43-yr-old 
woman diagnosed with profound mental retardation who was nonambulatory. She 
frequently engaged in head hitting. Laura did not reliably follow instructions or use any 
recognizable means of communication. Robert was a 30-yr-old man diagnosed with 
profound mental retardation who caused injury by rubbing his arms together forcefully. 
Robert had difficulty walking and spent most of his time in a wheelchair. He followed a 
few simple one-step directions and used a few gestures to communicate. 

Laura's sessions during the first four phases were conducted at a day program 
located on the grounds of the state residential facility, and observations during the fifth 
phase were conducted at her residence. All of Robert's sessions were conducted at his 
residence. 

Response Measurement and Reliability 

The primary dependent variables were SIB and object manipulation. Laura's SIB 
was defined as forcefully striking her head with either of her hands. Robert's SIB was 
defined as forcefully rubbing one arm against the other. Object manipulation was defined 
as physical contact with (e.g., touching, holding) a leisure item. 

Data on Laura's behavior were collected on handheld computers during 
continuous 10-s intervals. Data on Robert 's behavior were collected using paper and 
pencil. Each data sheet had a column listing the 60 10-s intervals of a 10-min session. 



34 
There were two additional columns in which observers recorded frequency of SIB and 

the occurrence of item manipulation for each 10-s interval. Data for both participants 

were summarized as either responses per minute (SIB) or percentage of intervals during 

which responding occurred (object manipulation). 

Interobserver agreement was assessed by having a second observer independently 

collect data during 36.9% and 25.5% of Laura's and Robert's sessions, respectively. 

Observers' records were compared on an interval-by-interval basis. Agreement for data 

on SIB was calculated by dividing the smaller number of responses by the larger number 

of responses for each interval and averaging these values across the session. Agreement 

coefficients for object manipulation were calculated by dividing the number of intervals 

containing agreements by the total number of intervals and multiplying by 100%. An 

interval was considered an agreement if both observers scored either the presence or 

absence of behavior. Mean agreement scores were 95.2% (range, 75.7% to 100%) and 

92.8% (range, 83.3% to 100%), respectively, for Laura's and Robert's SIB; and 91.7% 

(range, 47.4% to 100%) and 93.7% (range, 81.7% to 100%), respectively, for Laura's and 

Robert's object manipulation. 

Phase 1 : Functional Analysis 

During Phase 1, participants were exposed to four assessment conditions (alone, 
attention, demand, and play) in a multielement functional analysis based on procedures 
described by Iwata, Dorsey, Slifer, Bauman, and Richman (1982/1994). During the 
attention condition, the participant had access to leisure materials, and the experimenter 
ignored the participant except to express concern each time the participant engaged in 
SIB. This condition was a test for behavioral sensitivity to positive reinforcement in the 



35 
form of attention. During the demand condition, the experimenter initiated instructional 

trials on a fixed-time (FT) 30-s schedule using a series of graduated prompts and allowed 

the participant to escape the trial contingent on SIB. This condition was a test for 

behavioral sensitivity to negative reinforcement in the form of escape from demands. In 

the alone condition, the participant did not have access to leisure materials, and no social 

consequences were placed on SIB. This condition was designed to determine whether 

SIB persisted in the absence of social consequences. During the play condition, the 

participant had access to leisure materials, and the experimenter delivered attention to the 

participant on an FT 30-s schedule. This condition was a control for the other test 

conditions. 

Figure 4 shows rates (responses per min) of SIB exhibited by Laura and Robert 

during the functional analysis. Laura (top panel) engaged in SIB across all conditions (M 

= 5.1), but her highest rates of SIB occurred during the alone condition (M = 8.4). Robert 

(bottom panel) also engaged in SIB in all conditions (M = 2.5). Although his data 

contained a number of overlapping points, Robert's highest overall rates of SIB occurred 

during the alone condition (M = 3.5). Laura's and Robert's SIB also persisted during the 

alone sessions conducted at the conclusion of their functional analyses. These results 

suggest that Laura's and Robert's SIB was maintained by automatic reinforcement. 

Phase 2: Preference Assessment 

The participants' preference for a number of leisure items was assessed using 
procedures similar to those described by DeLeon, Iwata, Conners, and Wallace (1999). 
Before the assessment, the therapist familiarized the participants with each 



36 



C/3 
<U 
VI 

a 

o 

ex 

s 

OQ 

t— < 

C/3 



18 n 

16 

14 

12- 

10- 
8- 
6- 
4- 
2- 




Laura 




8-, Robert 



6- 



4- 



2- 



0- 



T" 

5 



10 



15 



10 



'T- 
IS 



- 1 - 

20 



- r - 
20 



25 




— l 
25 



Sessions 



Figure 4: Rates of SIB for Laura and Robert across functional analysis conditions 



37 
leisure item. Then, the individuals were allowed access to one leisure item at a time for 4 

min while an observer recorded duration of item contact (using a stopwatch) and rate of 

SIB. If the participant dropped the item during the assessment, the item was retrieved 

and placed within the participant's reach but was not placed in the individual's hands. 

Each item was assessed on three different occasions, for a total of 12 min per item. Ten 

items were assessed for each participant. 

Figure 5 shows results of the preference assessment. Each leisure item is 
represented by two data points: The squares show cumulative duration of item contact, 
scaled on the left-Y axis; the triangles represent rates of SIB, scaled on the right-Y axis. 
The points plotted are means from the three, 4-min assessment periods conducted for 
each item and are organized in descending order from left to right based on mean 
duration of item contact. There is an inverse relationship between object manipulation 
and SIB for both participants~this is especially true of Robert's data (bottom panel). 

The item labeled "C" for each participant was associated with the longest duration 
of item contact and the lowest rate of SIB. This item was delivered during NCR constant 
sessions. The item identified for Laura was a ribbon, which she manipulated with one 
hand continuously throughout the assessment. (Her mean rate of SIB when she had the 
ribbon was 0.2 responses per min.) The item selected for Robert was a bumble ball, 
which he turned on and manipulated with both hands continuously throughout the 
assessment. (He did not engage in SIB while he had access to the item.) 

The items labeled "V" were other leisure materials that were also associated with 
long durations of item contact and low rates of SIB. These items were delivered during 
NCR varied sessions. The items identified for Laura were a string of beads, a plastic ring 



38 



o 

| 
o 

2 
3 

Q 

<i> 

_> 

"■+3 

1 

3 





Laura 



4n 



3- 



2- 



I 
O 



E c 



Item Contact 



SIB 



C V 



Robert 



V 



4-1 



3- 



2- 



1- 



V V 



rlO 



- 8 



- 6 



- 4 



- 2 



U 

1/5 

d 

o 

6 m 

00 



-5 

-4 

3 

2 

1 



Stimuli 



Figure 5: Leisure item preference assessment results for Laura and Robert 



39 

with string, and a rubber worm. The items identified for Robert were a radio, a vibrating 

switch, and a hand-held massager. 

Phase 3: Brief NCR Evaluation 

The effects of NCR were evaluated during 10-min sessions using multiple 
baseline and reversal designs. The baseline condition was identical to the alone condition 
of the functional analysis: The participant did not have access to leisure items or to social 
interaction. During NCR, the individual had free access to the leisure item associated 
with the longest duration of item contact and the lowest rate of SIB during the preference 
assessment. At the beginning of each session, the leisure item was placed on the 
participant's wheelchair tray (Laura) or on a small table within the individual's reach 
(Robert). No instructions were delivered. If the participant dropped the item during the 
session, the item was retrieved and placed within the participant's reach but was not 
placed in the individual's hands. The therapist did not provide any other form of social 
interaction during the session. At the end of the session, the participant was informed 
that the session was over, and the materials were removed. 

Figure 6 shows responses per min of SIB during 1 0-min baseline and NCR 
sessions for Laura and Robert. Laura (top panel) engaged in variable but often high rates 
of SIB during baseline (M = 8.3; range, to 25.8). Her rate of SIB immediately 
decreased when she was given access to the ribbon (M = 0.9; range, to 2.7). The mean 
percentage of intervals during which Laura manipulated the ribbon was 94.7% (range, 
82.5% to 100%). Robert (bottom panel) engaged in somewhat more stable, moderate 
rates of SIB during baseline (M = 3.3; range, 1.5 to 5.5). Robert's rate of SIB quickly 
decreased when he had access to the bumble ball (M = 0.3; range, to 1 .0). His mean 



40 



30-, 



BL NCR BL 



NCR 



20- 



10- 



& 
a 

GO 




Figure 6: Rates of SIB for Laura and Robert across 10-min baseline and NCR sessions 



41 
level of object manipulation was 96.2% (range, 88.3% to 100%). Thus, data for both 

individuals showed that continuous access to their most preferred leisure item 

immediately produced large decreases in SIB. 

Phase 4: Extended NCR Evaluation 

The effects of NCR were evaluated during 2-hr sessions using a reversal design. 
During the first condition, NCR constant, the individual had access to the most preferred 
leisure item. This condition was identical to the NCR condition in Phase 3. During the 
second condition, NCR varied, the individual had free access to varied leisure items. 
These items were selected because they were associated with high levels of object 
manipulation and low levels of SIB during the leisure item preference assessment (Phase 
2). At the beginning of the NCR varied sessions, all leisure items were placed on the 
participant's wheelchair tray (Laura) or on a nearby table (Robert), and the participant 
was allowed to manipulate any of the items throughout the session. 

Figure 7 shows rates of SIB across 120-min NCR constant and NCR varied 
sessions. Each data point represents a 20-min segment; each connected series of data 
points represents one, 120-min session. Laura's rate of SIB (top panel) was low at the 
beginning of her first NCR constant session, but it increased as the session continued. 
This effect can be seen in several other NCR constant sessions (M = 3.3; range, 0.1 to 
18.3). By contrast, SIB remained low throughout the 2-hr periods when she had access to 
varied leisure items (M = 0.4; range, to 4.6). An inverse relationship between Laura's 
levels of object manipulation and SIB was observed during some NCR constant sessions. 
By contrast, her levels of object manipulation remained high throughout the NCR varied 
condition. Thus, her mean level of object manipulation was lower during the NCR 



42 



20 -i 



Constant Varied Constant 



Varied 



t/3 

g 
o 
& 

u 

8 

M 

go 




20-Min Segments 



Figure 7: Results of 120-min NCR sessions for Laura and Robert 



43 
constant sessions (M= 58.3%, range, 16.7% to 98.3%) than during the NCR varied 

sessions (M » 95.2% ; range, 61.7% to 100%). According to anecdotal observations of 

Laura's NCR varied sessions, Laura usually manipulated one of the leisure items (the 

string of beads) to the exclusion of the other items. 

Robert's NCR constant sessions (bottom panel) also show increased response 

rates across time (M = 1.4; range, 0.1 to 2.8). However, when Robert had access to 

multiple leisure items, he continued to engage in high rates of SIB throughout the 120- 

min sessions (M = 2.3; range, 0.7 to 4.0). Robert's level of object manipulation was high 

at the beginning of the NCR constant sessions but quickly dropped off as the session 

progressed (M ■ 43.5; range, 5.0 to 100%). This pattern was not observed when he had 

access to multiple items; Robert engaged in low levels of object manipulation throughout 

the NCR varied sessions (M = 32.9; range, 10.0 to 76.7%). 

Phase 5: Naturalistic NCR Evaluation 

The effects of NCR were evaluated when the intervention was implemented all 
day. Because, during the previous phase, the therapeutic effects of NCR did not endure 
during 2-hr sessions for Robert, only Laura participated in this final NCR evaluation. 
During baseline, Laura was observed at home during unstructured activity times. 
(Typically, she did not have access to many leisure items during these periods.) Next, 
Laura was given access to varied leisure items throughout the day, and data were 
collected during 10-min periods when she had access to these items. Most of the 
observations were conducted between 9 a.m. and 12 p.m., but data were also collected 
periodically in the afternoon between 1 and 4 p.m. Observations were conducted during 
both indoor and outdoor group leisure periods and while Laura was alone in her room 



44 
depending on where Laura was when the observer visited her home. Data were collected 

three to five times per week for 5 months, and follow-up data were collected 4 and 7 

months later. 

Figure 8 shows Laura's rates of SIB across 10-min observation periods at her 

home. Her SIB was extremely variable during baseline (M = 6.2; range, to 23.6). 

During NCR, when Laura had free access to varied leisure items, SIB remained generally 

low, although periodic increases were observed during some sessions (M = 1.0; range, 

to 5.6). Laura engaged in high rates of object manipulation during the NCR varied 

condition (M = 93.1%; range, 70.7% to 100%). When follow-up observations were 

conducted four and seven months later, Laura's rate of SIB was .5 on both occasions, and 

her levels of object manipulation were 98.4% and 91.7%, respectively. 

Discussion 

Results from the present study indicated that reinforcers obtained through leisure 
item manipulation competed with the reinforcers obtained automatically through SIB 
during brief (10-min) NCR sessions. However, Laura's and Robert's SIB increased 
during 120-min sessions. This may have occurred because both participants showed 
satiation to the reinforcers obtained through object manipulation during the extended 
sessions, as reflected by decreased levels of object manipulation. 

Data from Laura's 2-hr NCR varied sessions suggested that access to multiple 
reinforcers might mitigate the effects of satiation during long NCR sessions. When she 
had access to multiple, highly preferred leisure items, item manipulation remained high 
and SIB remained low throughout the 2-hr sessions. Laura's results followed the pattern 
of results obtained with Angela in the analog NCR experiment (Experiment 1). 



45 



Laura 



<D 

CO 

o 

Oh 

t/3 

0) 

g 

CQ 

C/3 



Baseline 



NCR Varied 



30 -i 

25- 

20- 

15 

10 

5 





i-lOO 




Figure 8: Rates of SIB for Laura across 10-min baseline and NCR 
observation periods at her home 



46 
It is interesting to note that, according to anecdotal observations of Laura's NCR 

varied sessions, she often manipulated the beads almost exclusively. Thus, if the beads 

had been available during the NCR constant sessions, it is possible that object 

manipulation may have remained high and SIB remained low throughout the sessions, as 

was observed during the NCR varied sessions. (Those results would have been similar to 

the pattern of results obtained with Matthew in Experiment 1.) The beads were not 

available during the NCR constant sessions because the results of the leisure item 

assessment suggested that the ribbon was equally preferred and was associated with 

lower levels of SIB. Nonetheless, the current results and the anecdotal reports from 

Laura's NCR evaluations suggest that providing multiple items may increase the chances 

that at least one of the items will effectively compete with the target behavior, even if 

reinforcer variation per se was not the key to extending the long-term therapeutic effects 

of NCR. 

Data from Laura's naturalistic NCR evaluation demonstrated that the intervention 
remained effective for several months when implemented every day during unstructured 
activity times. These results are promising because Laura's treatment gains were 
maintained with very little effort on the caregivers' part. 

The results from Robert's evaluation indicated that, although NCR was effective 
during 10-min sessions, the effects were not maintained during 120-min sessions. 
Providing varied leisure items to Robert seemed to have little or no effect on his 
behavior. Thus, for Robert, additional interventions would be necessary to achieve long- 
term reductions of SIB. 



GENERAL DISCUSSION 

The current experiments examined the short- versus long-term effects of NCR on 
behavior problems maintained by automatic reinforcement. In the first experiment, 
manipulation of two differentially preferred leisure items was studied because that 
behavior was considered analogous to the response options available when NCR is used 
to treat SIB maintained by automatic reinforcement. Manipulating the low-preference 
item was likened to engaging in SIB, and manipulating the high-preference item was 
likened to manipulating the leisure item provided noncontingently during treatment. 

Results obtained during 10-min sessions indicated that, for both participants, the 
reinforcer obtained through manipulation of a high-preference item competed with the 
reinforcer obtained through manipulation of a low-preference item during brief sessions. 
These results were consistent with numerous studies in which NCR has been evaluated 
during brief sessions. Although these results were neither novel nor surprising, they were 
a precondition for examining the long-term effects of NCR. 

During the next phase, Matthew manipulated the high-preference item (to the 
exclusion of the low-preference item) throughout the 2-hr sessions. By contrast, Angela 
manipulated the high-preference item less often and manipulated the low-preference item 
more often as the 2-hr sessions progressed. Matthew's data suggest that, for at least some 
individuals, NCR may remain effective for long periods of time. However, the results 
from Angela's evaluation suggest that NCR may lose its effectiveness when used for long 

47 



48 
periods of time because the individual may become satiated to the reinforcer obtained 

through item manipulation. Even in cases in which NCR loses its effectiveness over 

time, Angela's data demonstrated that the effects of satiation may be ameliorated when 

multiple high-preference items are available. 

The discrepancy between Matthew's and Angela's results may have been caused 
by the fact that the low-preference item selected for Matthew was rarely chosen during 
the leisure item preference assessment, whereas the low-preference item selected for 
Angela was chosen more often. Even though both participants continuously manipulated 
the items when they were presented singly, the difference in the rankings of the two low- 
preference items may have determined whether the participant would manipulate the 
object when only one other leisure item (the high-preference item) was available during 
the extended sessions. 

The difference between the method used to select Matthew's and Angela's low- 
preference leisure items and the discrepant outcomes that resulted may also shed light on 
why NCR may be effective for some individuals who engage in SIB but not for others. 
Treatment may fail, as it did for Angela, if the individual reverts back to SIB after short 
periods of leisure item manipulation. This outcome may be a function of the relative 
preference for SIB and manipulation of the leisure item available during NCR. NCR may 
quickly lose its effectiveness when leisure item manipulation is preferred only slightly 
more than SIB. 

The purpose of the second experiment was to examine the long-term effects of 
NCR on the SIB of two individuals, Laura and Robert. Results of functional analyses 
indicated that both individuals' target behaviors were maintained by automatic 



49 
reinforcement. After assessments were conducted to identify leisure items that competed 

with SIB, the effects of NCR were evaluated during 10-min sessions. Treatment effects 

were then assessed during 2-hr sessions when the participants had access to either (a) 

their most preferred leisure item, or (b) varied leisure items. Finally, the long-term 

effects of implementing NCR under naturalistic conditions were assessed for Laura. 

Results of the second experiment indicated that reinforcers obtained through 
leisure item manipulation competed with the reinforcers obtained automatically through 
SIB during brief NCR sessions. However, data from Laura's and Robert's 2-hr sessions 
suggested a loss of treatment effectiveness over time, as satiation to the reinforcers 
obtained through object manipulation occurred. Additional results obtained with Laura 
suggested that providing varied reinforcers might mitigate the effects of satiation during 
long NCR sessions. However, reinforcer variation seemed to have little or no effect on 
Robert's SIB. Data collected when NCR was implemented during Laura's daily 
unstructured activity times demonstrated that the procedure remained effective over 
several months. 

The difference between Laura's and Robert's results during the varied NCR 
condition was unexpected. NCR had similar effects during the brief NCR evaluation and 
during the extended NCR constant sessions, and similar results were expected during the 
NCR varied condition. The discrepancy was also unexpected because both participants 
had a limited repertoire of leisure skills. Laura and Robert manipulated all of their 
preferred items in a consistent manner: Laura twirled preferred items in the air, and 
Robert held vibrating items against his head and chest. The only item Robert 
manipulated that did not noticeably vibrate was the radio. However, Robert turned up the 



50 

volume and placed the radio against his ear, producing a mild form of vibration against 

his head and eardrum. Thus, both participants seemed to produce similar sensory 
consequences regardless of which preferred leisure item was available. Despite these 
similarities between Laura's and Robert's behavior, providing varied leisure items during 
extended NCR sessions had different effects. This discrepancy highlights the importance 
of evaluating the short- and long-term effects of a treatment for each individual for whom 
the intervention will be implemented. 

The results from the current experiments are relevant to previous findings in a 
number of ways. First, decreases in the target behavior were observed when leisure items 
were available during 10-min sessions. This finding replicates those from many other 
studies in which NCR has been shown to be effective during brief sessions (Bailey & 
Meyerson, 1970; Favell et al., 1982; Goh et al., 1995; Lockwood & Bourland, 1982; 
Piazza et al., 1998; Roscoe et al., 1998; Shore et al., 1997; Vollmer et al., 1994). 

The results of Robert's and Laura's leisure item preference assessments were 
consistent with research demonstrating that NCR is more effective when preferred leisure 
items are delivered noncontingently than when nonpreferred items are used (Vollmer et 
al., 1994). Because the purpose of the current study was not to demonstrate the 
difference between the effects of delivering preferred versus nonpreferred stimuli on self- 
injury, nonpreferred items were not delivered during the 10-min NCR sessions in the 
current experiment. Nonetheless, data from the brief leisure item probes conducted 
during Experiment 2 indicated that there was an inverse relationship between leisure item 
preference (as measured by duration of item contact) and SIB. 



51 
The leisure item assessment results were somewhat inconsistent with research 

suggesting that it may be beneficial to present stimuli that appear to provide stimulation 

similar to that obtained through the target behavior (Favell et al., 1982; Goh et al., 1995; 

Lockwood & Bourland, 1982; Piazza et al., 1998; Shore et al., 1997). The leisure items 

associated with the longest duration of item contact and lowest level of SIB in 

Experiment 2 did not appear to produce sensory stimulation similar to that obtained by 

engaging in SIB. The most preferred and effective leisure items identified for Laura were 

objects that could be twirled in the air by holding on to one end and rapidly rotating her 

wrist. The stimulation produced by twirling these items did not appear to match the 

stimulation Laura produced by hitting her head. The items associated with high duration 

of item contact and low levels of SIB for Robert were items that vibrated. He held these 

items against his face, head, and chest. The stimulation he produced when he 

manipulated these items had no apparent similarity to the stimulation he produced by 

forcefully rubbing his arms together. 

Laura's and Robert's preference for stimuli that did not produce stimulation that 

matched that produced by SIB was surprising in light of previous research. For example, 

Piazza et al. (1998) conducted a preference assessment to determine whether stimuli that 

produced oral stimulation (matched stimuli) were preferred over stimuli that produced 

other forms of stimulation (non-matched stimuli) by three individuals who engaged in 

pica. All three participants preferred matched stimuli, and noncontingent access to 

matched stimuli was associated with lower levels of pica than access to non-matched 

stimuli. 



52 
The benefit of delivering stimuli that produced consequences similar to those 

produced by the target behavior may not have been observed in the current study because 

stimuli that met this criterion were not included in the leisure item assessment. Items 

were selected for the assessment because they were either reported to be preferred by the 

participants or because they produced a variety of sensory consequences. Items were not 

selected because they appeared to produce similar sensory consequences to those 

produced by the target behavior. Thus, if different criteria were used to select leisure 

items for the initial assessment, we may have found that the stimuli associated with high 

levels of contact and low levels of SIB did produce sensory consequences similar to those 

produced by SIB. 

The results from Angela's (Exp 1) and Laura's (Exp 2) extended NCR sessions 

were consistent with research demonstrating that stimulus variation may improve the 

effectiveness of reinforcement. During Angela's extended NCR constant condition, she 

manipulated the high-preference item less and manipulated the medium-preference item 

more as the session progressed. During Laura's extended NCR constant condition, she 

manipulated the leisure item less and engaged in SIB more toward the end of the 2-hr 

sessions. These results were not observed for either participant during the extended NCR 

varied condition when varied high-preference leisure items were available: Both 

participants manipulated the high-preference items throughout the 2-hr sessions. These 

results are similar to results obtained by Egel (1981), who found that correct responding 

and on-task behavior showed declining trends within sessions when the same reinforcer 

was consistently presented to three participants. By contrast, stable levels of correct 



53 
responding and on-task behavior were observed in all three participants when varied 

reinforcers were presented. 

Angela's and Laura's data from the extended NCR evaluation and Laura's data 
from the naturalistic NCR evaluation are also consistent with results demonstrating that 
NCR, when used as part of a treatment package, can be effective during extended periods 
and under naturalistic conditions (Vollmer et al., 1994). Vollmer et al. trained the 
families of two participants to implement treatment packages that included an NCR 
procedure for one hour per day. Data were collected periodically to evaluate the 
effectiveness of the intervention package under these naturalistic conditions, and results 
indicated that the intervention was effective for both participants. Similar results were 
obtained in the current investigation when Angela and Laura participated in 2-hr NCR 
sessions, and when Laura's caregivers were trained to implement NCR at her home 
during daily unstructured time periods. 

The results from Robert's extended NCR sessions indicated that NCR was 
ineffective as an intervention when implemented without additional treatments. This 
finding is consistent with results obtained by Vollmer et al. (1994). After implementing 
NCR to decrease three participants' hand mouthing, the experimenters reinforced the 
object manipulation of two participants to obtain further decreases in SIB. A manual 
restraint procedure was also necessary to decrease one of the participant's SIB to 
therapeutic levels. 

The current results are also relevant to previous research conducted on reinforcer 
substitutability. Substitutable reinforcers have been defined as stimuli "for which an 
increase in consumption of one alternative results in decreased consumption of its 



54 
substitute" (Shore et al., 1997, p. 23) Shore et al. conducted probes to identify leisure 

items that appeared to produce reinforcers substitutable with SIB that was maintained by 

automatic reinforcement. Items that were associated with the highest levels of item 

manipulation and lowest levels of SIB were considered substitutable with SIB. A similar 

method was used in Experiment 2 of the current investigation to identify leisure items to 

be delivered during NCR. During the brief NCR evaluation, the reinforcers obtained 

through manipulating high-preference leisure items appeared to be substitutable with the 

reinforcers obtained through engaging in SIB. However, these effects were not 

maintained during extended NCR sessions. 

The results of the extended NCR constant condition suggest a need to investigate 

and perhaps define substitutability more clearly. It may be useful to investigate what 

effects altering the establishing operation of one reinforcer have on the establishing 

operation of the other. For example, when two responses that produce different 

reinforcers are available concurrently, an individual may engage in one response more 

than the other. However, once the individual becomes satiated to the reinforcer 

maintaining the first behavior, the second behavior may increase. Even though 

consumption of one reinforcer initially resulted in decreased consumption of the other, 

this pattern of behavior would suggest that the two reinforcers maintaining the behaviors 

were independent. If the two reinforcers were truly substitutable, it seems that satiation 

to one reinforcer would produce satiation to the second reinforcer, as evidenced by 

temporary decreases in both behaviors after an extended time in which the individual 

emitted the first response. 



55 
According to these criteria for considering two reinforcers substitutable, the 

reinforcer Laura and Robert obtained by manipulating their most preferred leisure item 

competed with the reinforcer maintaining SIB, but the two reinforcers were not 

substitutable. Likewise, the reinforcers Angela obtained by manipulating her low- and 

high-preference leisure items did not appear to be substitutable. The relationship 

between the reinforcers obtained by Matthew is unclear because he never appeared to 

satiate to the high-preference leisure item. 

The current study contained a few noteworthy limitations. First, data were not 
collected to identify which leisure items were being manipulated during the extended 
varied NCR sessions. According to anecdotal reports, Laura manipulated the beads to the 
exclusion of the other leisure items during this condition. Collecting data would have 
aided in determining the function providing varied items had on her SIB. Stimulus 
variation may have decreased the likelihood that Laura would satiate to any one 
reinforcer. On the other hand, providing varied items may have merely enabled her to 
select the one item to which she would not quickly satiate. In either case, stimulus 
variation effectively extended the treatment utility of NCR during 2-hr sessions and when 
implemented under naturalistic conditions over several months. 

Second, no additional interventions were attempted for Robert. Due to parental 
and staff concerns regarding Robert's health, he was referred only for participation in the 
current investigation. Evaluation of other treatments was considered beyond the scope of 
the current investigation. Nonetheless, evaluating interventions designed to increase the 
variety of Robert's leisure skills and to decrease his SIB would have been advisable. 



56 
Despite these limitations, this study highlights the importance of evaluating 

interventions under naturalistic conditions as well as under more tightly-controlled 

conditions. Initial treatment evaluations and refinements need to be conducted under 

well-controlled conditions. However, treatments should eventually be tested under 

conditions similar to those in which the treatment will be used because different results 

may be obtained when a treatment is evaluated under naturalistic conditions than under 

more limited conditions. Investigators should examine treatment generality and identify 

procedures by which generality can be achieved. 

Future research should also concentrate on increasing leisure skills among 

individuals for whom NCR is ineffective. NCR fails to reduce problem behaviors 

maintained by automatic reinforcement during even brief sessions in individuals who do 

not have adequate leisure skills. Improving leisure skills may also prove beneficial for 

individuals who engage in a limited repertoire of leisure item manipulation and for whom 

NCR is only effective during brief sessions. Providing varied stimuli during NCR to 

extend the therapeutic effects of the intervention is more likely to be effective for 

individuals who manipulate a variety of leisure items. 



REFERENCES 



Bailey, J., & Meyerson, L. (1970). Effect of vibratory stimulation on a retardate's 
self- injurious behavior. Psychological Aspects of Disability. 17 . 133-137. 

Cowdery, G. E., Iwata, B. A., & Pace, G. M. (1990). Effects and side effects of 
DRO as treatment for self-injurious behavior. Journal of Applied Behavior Analysis. 23 . 
497-506. 

DeLeon, I. G., Iwata, B. A., Conners, J., & Wallace, M. D. (1999). Examination 
of ambiguous stimulus preferences with duration-based measures. Journal of Applied 
Behavior Analysis. 32 . 111-1 14. 

Dorsey, M. F., Iwata, B. A., Reid, D. H., & Davis, P. A. (1982). Protective 
equipment: Continuous and contingent application in the treatment of self-injurious 
behavior. Journal of Applied Behavior Analysis. 15 . 217-230. 

Egel, A. L. (1981). Reinforcer variation: Implications for motivating 
developmentally disabled children. Journal of Applied Behavior Analysis. 14 . 345-350. 

Favell, J. E., McGimsey, J. F., & Schell, R. M. (1982). Treatment of self-injury by 
providing alternate sensory activities. Analysis and Intervention in Developmental 
Disabilities. 2 . 83-104. 

Fischer, S. M., Iwata, B. A., & Mazaleski, J. L. (1997). Noncontingent delivery of 
arbitrary reinforcers as treatemnt for self-injurious behavior. Journal of Applied Behavior 
Analysis. 30 . 239-249. 

Fisher, W. W., Lindauer, S. E., Alterson, C. J., & Thompson, R. H. (1998). 
Assessment and treatment of destructive behavior maintained by stereotypic 
manipulation. Journal of Applied Behavior Analysis. 31 . 513-527. 

Goh, H., Iwata, B. A., Shore, B. A., DeLeon, I. G, Lerman, D. C, Ulrich, S. M., 
& Smith, R. G. (1995). An analysis of the reinforcing properties of hand mouthing. 
Journal of Applied Behavior Analysis. 28 . 269-283. 

Hagopian, L. P., Fisher, W. W., & Legacy, S. M. (1994). Schedule effects of 
noncontingent reinforcement on attention-maintained destructive behavior. Journal of 
Applied Behavior Analysis. 27 . 317-325. 

57 



58 

Iwata, B. A., Dorsey, M. F., Slifer, K. J., Bauman, K. E., & Richman, G. S. 
(1994). Toward a functional analysis of self-injury. Journal of Applied Behavior 
Analysis. 27 . 197-209. (Reprinted from Analysis and Intervention in Developmental 
Disabilities. 2. 3-20. 1982) 

Iwata, B. A., Pace, G. M., Dorsey, M. F., Zarcone, J. R., Vollmer, T. R., Smith, R. 
G, Rodgers, T. A., Lerman, D. C, Shore, B. A., Mazaleski, J. L., Goh, H., Cowdery, G. 
E., Kalsher, M. J., McCosh, K. C., & Willis, K. D. (1994). The functions of self-injurious 
behavior: An experimental-epidemiological analysis. Journal of Applied Behavior 
Analysis. 27 . 215-240. 

Kennedy, C. H., & Souza, G (1995). Functional analysis and treatment of eye 
poking. Journal of Applied Behavior Analysis. 28 . 27-37. 

Lalli, J. S., Livezey, K., & Kates, K. (1996). Functional analysis and treatment of 
eye poking with response blocking. Journal of Applied Behavior Analysis. 29 . 129-132. 

Lindberg, J. S., Iwata, B. A., & Kahng, S. (1999). On the relation between object 
manipulation and stereotypic self-injurious behavior. Journal of Applied Behavior 
Analysis. 32 . 51-62. 

Lockwood, K, & Bourland, G. (1982). Reduction of self-injurious behaviors by 
reinforcement and toy use. Mental Retardation. 20 . 1 69- 1 73 . 

Mace, F. C, & Lalli, J. S. (1991). Linking descriptive and experimental analyses 
in the treatment of bizarre speech. Journal of Applied Behavior Analysis. 24 . 553-562. 

Mazaleski, J. L., Iwata, B. A., Rodgers, T. A., Vollmer, T. R., & Zarcone, J. R. 
(1994). Protective equipment as treatment for stereotypic hand mouthing: Sensory 
extinction or punishment effects? Journal of Applied Behavior Analysis. 27 . 345-355. 

Michael, J. (1993). Establishing operations. The Behavior Analyst. 16 . 191-206. 

Parrish, J. M., Aguerrevere, L., Dorsey, M. F., & Iwata, B. A. (1980). The effects 
of protective equipment on self-injurious behavior. The Behavior Therapist. 3 . 28-29. 

Piazza, C. C., Fisher, W. W., Hanley, G. P., LeBlanc, L. A., Worsdell, A. S., 
Lindauer, S. E., & Keeney, K. M. (1998). Treatment of pica through multiple analyses of 
its reinforcing functions. Journal of Applied Behavior Analysis. 31 . 165-189. 

Poling, A., & Normand, M. (1999). Noncontingent reinforcement: An 
inappropriate description of time-based schedules that reduce behavior. Journal of 
Applied Behavior Analysis. 32 . 237-238. 



59 

Rapp, J. T., Miltenberger, R. G., Galensky, T. L., Ellingson, S. A., & Long, E. S. 
(1999). A functional analysis of hair pulling. Journal of Applied Behavior Analysis. 32 . 
329-337. 

Rescorla, R. A., & Skucy, J. C. (1969). Effect of response-independent reinforcers 
during extinction. Journal of Comparative and Physiological Psychology. 67 . 381-389. 

Rincover, A. (1978). Sensory extinction: A procedure for eliminating self- 
stimulatory behavior in developmentally disabled children. Journal of Abnormal Child 
Psychology. 6 . 299-310. 

Ringdahl, J. E., Vollmer, T. R., Marcus, B. A., & Roane, H. S. (1997). An 
analogue evaluation of enviornmental enrichment: The role of stimulus preference. 
Journal of Applied Behavior Analysis. 30 . 203-216. 

Roscoe, E. M., Iwata, B. A., & Goh, H. (1998). A comparison of noncontingent 
reinforcement and sensory extinction as treatments for self-injurious behavior. Journal of 
A pplied Behavior Analysis. 3 1 . 635-646. 

Shore, B. A., & Iwata, B. A. (1999). Assessment and treatment of behavior 
disorders maintained by nonsocial (automatic) reinforcement. In A. C. Repp & R. H. 
Horner (Eds.), Functional analysis of problem behavior: From effective assessment to 
effective supp ort (pp. 1 17-146). Belmont, CA: Wadsworth. 

Shore, B. A., Iwata, B. A., DeLeon, I. G., Kahng, S., & Smith, R. G. (1997). An 
analysis of reinforcer substitutability using object manipulation and self-injury as 
competing responses. Journal of Applied Behavior Analysis. 30 . 21-41. 

Skinner, B. F. (1953). Science and human behavior . New York: Macmillan. 
Van Houten, R. (1993). The use of wrist weights to reduce self-injury maintained 
by sensory reinforcement. Journal of Applied Behavior Analysis. 26 . 197-203. 

Vaughan, M. E., & Michael, J. (1982). Automatic reinforcement: An important 
but ignored concept. Behaviorism. 10 . 217-227. 

Vollmer, T. R. (1994). The concept of automatic reinforcement: Implications for 
behavioral research in developmental disabilities. Research in Developmental 
Disabilities. 15 . 187-207. 

Vollmer, T. R. (1999). Noncontingent reinforcement: Some additional comments. 
Journal of Applied Behavior Analysis. 32 . 239-240. 

Vollmer, T. R., Iwata, B. A., Zarcone, J. R, Smith, R. G., & Mazaleski, J. L. 
(1993). The role of attention in the treatment of attention-maintained self-injurious 
behavior: Noncontingent reinforcement and differential reinforcement of other behavior. 
Journal of Applied Behavior Analysis. 26 . 9-21. 



60 

Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Treatment of self-injury and 
hand mouthing following inconclusive functional analyses. Journal of Applied Behavior 
Analysis. 27 . 331-344. 

Vollmer, T. R., Marcus, B. A., & Ringdahl, J. E. (1995a). Noncontingent escape 
as treatment for self-injurious behavior maintained by negative reinforcement. Journal of 
Applied Behavior Analysis. 28 . 15-26. 

Vollmer, T. R., Marcus, B. A., Ringdahl, J. E., & Roane, H. S. (1995b). 
Progressing from brief assessments to extended experimental analyses in the evaluation 
of aberrant behavior. Journal of Applied Behavior Analysis. 28 . 561-576. 

Wacker, D. P., Steege, M. W., Northup, J., Sasso, G., Berg, W., Reimers, T., 
Cooper, L., Cigrand, K., & Donn, L. (1990). A component analysis of functional 
communication training across three topographies of severe behavior problems. Journal 
of Applied Behavior Analysis. 23 . 417-429. 

Zhou, L., Goff, G. A., & Iwata, B. A. (2000). Effects of increased response effort 
on self-injury and object manipulation as competing responses. Journal of Applied 
Behavior Analysis. 33 . 29-40. 



BIOGRAPHICAL SKETCH 

Jana Lindberg received a Bachelor of Arts degree from Brigham Young 
University in 1995. Her major was an interdisciplinary program with a combined focus 
in organizational behavior, communications, and psychology. Jana's interest in applied 
behavior analysis began when she took an introductory course in the field as an 
undergraduate student. 

In the Fall of 1995, Jana began graduate studies in experimental and applied 
behavior analysis at the University of Florida. Her studies have included both theoretical 
and applied behavior analysis courses. She has specialized in the assessment and 
treatment of severe behavior and learning disorders. 

During Jana's five years of graduate school, she has worked as a research 
assistant under the direction of Dr. Brian Iwata at the Florida Center on Self-Injury. In 
this capacity, she has conducted assessment and treatment sessions, analyzed data, trained 
staff, served as the lab coordinator, supervised research, and disseminated results through 
publishing articles and giving conference presentations. In addition, she has worked as a 
teaching assistant and instructor at the University of Florida. 



61 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



S&-A 




Brian A. Iwata, Chairman 
Professor of Psychology 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 




Marc N. Branch 
Professor of Psychology 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 





Shari Ellis 

Assistant Professor of Psychology 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 

2*4 a £ UU^ 



Timothy R. Vollmer 

Assistant Professor of Psychology 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 

0[y^ 



Jennifer Asmus 

Assistant Professor of Educational 
Psychology 



This dissertation was submitted to the Graduate Faculty of the Department of 
Psychology in the College of Liberal Arts and Sciences and to the Graduate School and 
was accepted as partial fulfillment of the requirements for the degree of Doctor of 
Philosophy. 

May 2000 



Dean, Graduate School 



2000 












UNIVERSITY OF FLORIDA 



3 1262 08555 0282